Introduction

FIFA 22 is a football simulation video game published by Electronic Arts as part of the FIFA series. It is the 29th installment in the FIFA sereis, and was released worldwide on 1 October 2021 for Microsoft Windows, Nintendo Switch, PlayStation 4, PlayStation 5, Xbox One, and Xbox Series X/S.

source: Wikipedia

Set-up & Packages

library(tidyverse)
library(magrittr)
library(DataExplorer)
library(maps)
library(plotly)
library(DT)
library(tidytext)
library(gridExtra)
library(factoextra)
library(kableExtra)
library(splitstackshape)
library(ggthemes)
library(data.table)
library(waffle)
library(knitr)

options(scipen = 999)

DATA

df <- read.csv("../db/players_22.csv", encoding = "UTF-8")
kable(t(head(df,3)), "html") %>% kable_styling(bootstrap_options = c("striped","hover", "condensed"), font_size = 8) %>% scroll_box(width = "100%")

	1	2	3
sofifa_id	158023	188545	20801
player_url	https://sofifa.com/player/158023/lionel-messi/220002	https://sofifa.com/player/188545/robert-lewandowski/220002	https://sofifa.com/player/20801/c-ronaldo-dos-santos-aveiro/220002
short_name	L. Messi	R. Lewandowski	Cristiano Ronaldo
long_name	Lionel Andrés Messi Cuccittini	Robert Lewandowski	Cristiano Ronaldo dos Santos Aveiro
player_positions	RW, ST, CF	ST	ST, LW
overall	93	92	91
potential	93	92	91
value_eur	78000000	119500000	45000000
wage_eur	320000	270000	270000
age	34	32	36
dob	1987-06-24	1988-08-21	1985-02-05
height_cm	170	185	187
weight_kg	72	81	83
club_team_id	73	21	11
club_name	Paris Saint-Germain	FC Bayern München	Manchester United
league_name	French Ligue 1	German 1. Bundesliga	English Premier League
league_level	1	1	1
club_position	RW	ST	ST
club_jersey_number	30	9	7
club_loaned_from
club_joined	2021-08-10	2014-07-01	2021-08-27
club_contract_valid_until	2023	2023	2023
nationality_id	52	37	38
nationality_name	Argentina	Poland	Portugal
nation_team_id	1369	1353	1354
nation_position	RW	RS	ST
nation_jersey_number	10	9	7
preferred_foot	Left	Right	Right
weak_foot	4	4	4
skill_moves	4	4	5
international_reputation	5	5	5
work_rate	Medium/Low	High/Medium	High/Low
body_type	Unique	Unique	Unique
real_face	Yes	Yes	Yes
release_clause_eur	144300000	197200000	83300000
player_tags	#Dribbler, #Distance Shooter, #FK Specialist, #Acrobat, #Clinical Finisher, #Complete Forward	#Aerial Threat, #Distance Shooter, #Clinical Finisher, #Complete Forward	#Aerial Threat, #Dribbler, #Distance Shooter, #Crosser, #Acrobat, #Clinical Finisher, #Complete Forward
player_traits	Finesse Shot, Long Shot Taker (AI), Playmaker (AI), Outside Foot Shot, One Club Player, Chip Shot (AI), Technical Dribbler (AI)	Solid Player, Finesse Shot, Outside Foot Shot, Chip Shot (AI)	Power Free-Kick, Flair, Long Shot Taker (AI), Speed Dribbler (AI), Outside Foot Shot
pace	85	78	87
shooting	92	92	94
passing	91	79	80
dribbling	95	86	88
defending	34	44	34
physic	65	82	75
attacking_crossing	85	71	87
attacking_finishing	95	95	95
attacking_heading_accuracy	70	90	90
attacking_short_passing	91	85	80
attacking_volleys	88	89	86
skill_dribbling	96	85	88
skill_curve	93	79	81
skill_fk_accuracy	94	85	84
skill_long_passing	91	70	77
skill_ball_control	96	88	88
movement_acceleration	91	77	85
movement_sprint_speed	80	79	88
movement_agility	91	77	86
movement_reactions	94	93	94
movement_balance	95	82	74
power_shot_power	86	90	94
power_jumping	68	85	95
power_stamina	72	76	77
power_strength	69	86	77
power_long_shots	94	87	93
mentality_aggression	44	81	63
mentality_interceptions	40	49	29
mentality_positioning	93	95	95
mentality_vision	95	81	76
mentality_penalties	75	90	88
mentality_composure	96	88	95
defending_marking_awareness	20	35	24
defending_standing_tackle	35	42	32
defending_sliding_tackle	24	19	24
goalkeeping_diving	6	15	7
goalkeeping_handling	11	6	11
goalkeeping_kicking	15	12	15
goalkeeping_positioning	14	8	14
goalkeeping_reflexes	8	10	11
goalkeeping_speed	NA	NA	NA
ls	89+3	90+2	90+1
st	89+3	90+2	90+1
rs	89+3	90+2	90+1
lw	92	85	88
lf	93	88	89
cf	93	88	89
rf	93	88	89
rw	92	85	88
lam	93	86+3	86+3
cam	93	86+3	86+3
ram	93	86+3	86+3
lm	91+2	84+3	86+3
lcm	87+3	80+3	78+3
cm	87+3	80+3	78+3
rcm	87+3	80+3	78+3
rm	91+2	84+3	86+3
lwb	66+3	64+3	63+3
ldm	64+3	66+3	59+3
cdm	64+3	66+3	59+3
rdm	64+3	66+3	59+3
rwb	66+3	64+3	63+3
lb	61+3	61+3	60+3
lcb	50+3	60+3	53+3
cb	50+3	60+3	53+3
rcb	50+3	60+3	53+3
rb	61+3	61+3	60+3
gk	19+3	19+3	20+3
player_face_url	https://cdn.sofifa.net/players/158/023/22_120.png	https://cdn.sofifa.net/players/188/545/22_120.png	https://cdn.sofifa.net/players/020/801/22_120.png
club_logo_url	https://cdn.sofifa.net/teams/73/60.png	https://cdn.sofifa.net/teams/21/60.png	https://cdn.sofifa.net/teams/11/60.png
club_flag_url	https://cdn.sofifa.net/flags/fr.png	https://cdn.sofifa.net/flags/de.png	https://cdn.sofifa.net/flags/gb-eng.png
nation_logo_url	https://cdn.sofifa.net/teams/1369/60.png	https://cdn.sofifa.net/teams/1353/60.png	https://cdn.sofifa.net/teams/1354/60.png
nation_flag_url	https://cdn.sofifa.net/flags/ar.png	https://cdn.sofifa.net/flags/pl.png	https://cdn.sofifa.net/flags/pt.png

Ther are 110 columns and 19239 observations on the data.

dim(df)

## [1] 19239   110

Now we gonna check some more details about data with DataExplorer package.

# Check data structure with introduce()
kable(introduce(df), "html") %>% kable_styling(bootstrap_options = c("striped","hover", "condensed")) %>% scroll_box(width = "100%")

rows	columns	discrete_columns	continuous_columns	all_missing_columns	total_missing_values	complete_rows	total_observations	memory_usage
19239	110	50	60	0	68414	0	2116290	20813248

# Now we will plot it with plot_intro()
plot_intro(df, title = "Data Structure of database")

plot_missing(df, missing_only = TRUE, title = "Percentage of Missing Features")

Data Prep.

First, we need to deal with missing features in ours dataset.

Missing Values

Free agent player & retire player

Since free agent player didn’t play for any clubs. Therefore, some features (club_contract_valid_unitl, club_jersy_number, league_level, club_team_id, wage_eur, value_eur) will be null.

df %>% 
  filter(is.na(club_team_id) &
          is.na(club_jersey_number) &
          is.na(league_level) &
          is.na(club_team_id) &
          is.na(wage_eur) &
          is.na(value_eur)) %$%
  unique(club_name)

## [1] ""

However, we can see from graph above that there are some non-free agent player that didn’t have value. We also can assume that these old players value equal to 0.

df %>% 
  filter(is.na(value_eur) & club_name != "") %$%
  kable(head(.),"html") %>% 
  kable_styling(bootstrap_options = c("striped","hover", "condensed"), font_size = 8) %>% 
  scroll_box(width = "100%")

sofifa_id	player_url	short_name	long_name	player_positions	overall	potential	value_eur	wage_eur	age	dob	height_cm	weight_kg	club_team_id	club_name	league_name	league_level	club_position	club_jersey_number	club_joined	club_contract_valid_until	nationality_id	nationality_name	nation_team_id	nation_jersey_number	preferred_foot	weak_foot	skill_moves	international_reputation	work_rate	body_type	real_face	release_clause_eur	player_traits	pace	shooting	passing	dribbling	defending	physic	attacking_crossing	attacking_finishing	attacking_heading_accuracy	attacking_short_passing	attacking_volleys	skill_dribbling	skill_curve	skill_fk_accuracy	skill_long_passing	skill_ball_control	movement_acceleration	movement_sprint_speed	movement_agility	movement_reactions	movement_balance	power_shot_power	power_jumping	power_stamina	power_strength	power_long_shots	mentality_aggression	mentality_interceptions	mentality_positioning	mentality_vision	mentality_penalties	mentality_composure	defending_marking_awareness	defending_standing_tackle	defending_sliding_tackle	goalkeeping_diving	goalkeeping_handling	goalkeeping_kicking	goalkeeping_positioning	goalkeeping_reflexes	goalkeeping_speed	ls	st	rs	lw	lf	cf	rf	rw	lam	cam	ram	lm	lcm	cm	rcm	rm	lwb	ldm	cdm	rdm	rwb	lb	lcb	cb	rcb	rb	gk	player_face_url	club_logo_url	club_flag_url	nation_flag_url
138703	https://sofifa.com/player/138703/andres-dalessandro/220002	A. D’Alessandro	Andrés Nicolás D’Alessandro	CAM, RM	75	75	NA	900	40	1981-04-15	174	69	111325	Club Nacional de Football	Uruguayan Primera División	1	SUB	10	2021-01-04	2021	52	Argentina	NA	NA	Left	3	4	1	Medium/Low	Normal (170-185)	No	NA	Flair, Long Passer (AI), Playmaker (AI), Technical Dribbler (AI)	42	71	80	75	38	50	79	69	45	80	68	77	82	80	77	80	45	39	51	73	82	69	45	30	50	76	75	42	73	85	80	78	46	29	25	6	11	12	15	6	NA	66+2	66+2	66+2	72	72	72	72	72	75	75	75	71+2	71+2	71+2	71+2	71+2	53+2	58+2	58+2	58+2	53+2	49+2	47+2	47+2	47+2	49+2	17+2	https://cdn.sofifa.net/players/138/703/22_120.png	https://cdn.sofifa.net/teams/111325/60.png	https://cdn.sofifa.net/flags/uy.png	https://cdn.sofifa.net/flags/ar.png
152912	https://sofifa.com/player/152912/jose-sand/220002	J. Sand	José Gustavo Sand	ST	75	75	NA	13000	40	1980-07-17	182	79	110395	Club Atlético Lanús	Argentina Primera División	1	LS	9	2019-01-21	2022	52	Argentina	NA	NA	Right	4	3	1	High/Low	Normal (170-185)	No	NA	Solid Player, Finesse Shot, Team Player	51	79	62	68	37	70	53	81	77	69	77	67	63	63	46	75	49	53	38	81	67	77	71	41	83	73	73	33	83	69	85	85	45	21	25	11	16	15	9	14	NA	75	75	75	68	73	73	73	68	70+2	70+2	70+2	66+2	63+2	63+2	63+2	66+2	49+2	52+2	52+2	52+2	49+2	47+2	51+2	51+2	51+2	47+2	20+2	https://cdn.sofifa.net/players/152/912/22_120.png	https://cdn.sofifa.net/teams/110395/60.png	https://cdn.sofifa.net/flags/ar.png	https://cdn.sofifa.net/flags/ar.png
110381	https://sofifa.com/player/110381/maximiliano-rodriguez/220002	M. Rodríguez	Maximiliano Rubén Rodríguez	LW, LM, ST	73	73	NA	10000	40	1981-01-02	173	74	110396	Newell’s Old Boys	Argentina Primera División	1	SUB	11	2019-01-21	2021	52	Argentina	NA	NA	Right	4	4	3	Medium/Low	Normal (170-185)	No	NA	Leadership, Team Player	64	74	76	75	44	64	72	73	61	78	71	75	76	71	75	76	64	64	72	72	70	77	67	46	70	75	73	47	76	76	77	86	57	30	26	15	11	15	9	9	NA	72+1	72+1	72+1	73	74-1	74-1	74-1	73	75-2	75-2	75-2	72+1	70+3	70+3	70+3	72+1	57+3	60+3	60+3	60+3	57+3	54+3	54+3	54+3	54+3	54+3	18+3	https://cdn.sofifa.net/players/110/381/22_120.png	https://cdn.sofifa.net/teams/110396/60.png	https://cdn.sofifa.net/flags/ar.png	https://cdn.sofifa.net/flags/ar.png
115909	https://sofifa.com/player/115909/ruben-castro-martin/220002	Rubén Castro	Rubén Castro Martín	ST	69	69	NA	3000	40	1981-06-27	169	68	100851	FC Cartagena	Spanish Segunda División	2	ST	7	2020-09-12	2022	45	Spain	NA	NA	Right	3	3	2	Medium/Low	Normal (170-)	No	NA		70	70	61	72	29	61	60	71	62	60	66	70	68	67	45	69	72	69	86	75	81	69	85	62	61	68	54	37	77	74	75	68	36	12	14	14	8	15	8	12	NA	69	69	69	70-1	71-2	71-2	71-2	70-1	69	69	69	67+2	61+2	61+2	61+2	67+2	50+2	47+2	47+2	47+2	50+2	47+2	43+2	43+2	43+2	47+2	18+2	https://cdn.sofifa.net/players/115/909/22_120.png	https://cdn.sofifa.net/teams/100851/60.png	https://cdn.sofifa.net/flags/es.png	https://cdn.sofifa.net/flags/es.png
153066	https://sofifa.com/player/153066/lucas-licht/220002	L. Licht	Lucas Matías Licht	LB, LWB, LM	69	69	NA	5000	40	1981-04-06	174	72	101084	Gimnasia y Esgrima La Plata	Argentina Primera División	1	SUB	25	2012-07-21	2021	52	Argentina	NA	NA	Left	4	3	1	Medium/Medium	Normal (170-185)	No	NA	Leadership, Early Crosser, Team Player	64	61	65	72	68	66	80	52	53	52	57	76	73	68	71	69	59	68	69	66	73	71	72	64	66	68	69	71	56	64	84	72	69	71	69	8	14	6	15	8	NA	61+2	61+2	61+2	65	63	63	63	65	64+2	64+2	64+2	66+2	65+2	65+2	65+2	66+2	68+1	67+2	67+2	67+2	68+1	67+2	67+2	67+2	67+2	67+2	17+2	https://cdn.sofifa.net/players/153/066/22_120.png	https://cdn.sofifa.net/teams/101084/60.png	https://cdn.sofifa.net/flags/ar.png	https://cdn.sofifa.net/flags/ar.png
124344	https://sofifa.com/player/124344/ibrahim-ozturk/220002	İ. Öztürk	İbrahim Öztürk	CB	67	67	NA	3000	40	1981-06-21	186	80	101006	Altay SK	Turkish Süper Lig	1	CB	38	2017-08-24	2022	48	Turkey	NA	NA	Right	2	2	1	Low/High	Normal (185+)	No	NA	Dives Into Tackles (AI), Team Player	32	43	48	48	68	74	44	30	72	56	52	46	44	48	42	54	30	34	32	68	38	66	59	65	78	46	78	66	34	44	51	51	70	68	62	15	13	10	10	6	NA	50+2	50+2	50+2	44	47	47	47	44	46+2	46+2	46+2	46+2	52+2	52+2	52+2	46+2	57+2	62+2	62+2	62+2	57+2	59+2	67	67	67	59+2	17+2	https://cdn.sofifa.net/players/124/344/22_120.png	https://cdn.sofifa.net/teams/101006/60.png	https://cdn.sofifa.net/flags/tr.png	https://cdn.sofifa.net/flags/tr.png

Fill the missing values

# Fill the missing values
df$club_contract_valid_until[is.na(df$club_contract_valid_until)] <- 0
df$club_jersey_number[is.na(df$club_jersey_number)] <- 0

# league_level is ordinal variable which 1 is the highest league and 5 is lowest league
# Since, these player are free agent and didn't play in any league at the moment so we assign 6 to them
df$league_level[is.na(df$league_level)] <- 6

# club_team_id run from 1 to 115820. we will assing 0 to the free agent
df$club_team_id[is.na(df$club_team_id)] <- 0

df$wage_eur[is.na(df$wage_eur)] <- 0
df$value_eur[is.na(df$value_eur)] <- 0

plot_missing(df, missing_only = TRUE, title = "Percentage of Missing Features")

Release Clause

Not all player have release clause. So, it natural to have some missng release_clause_eur.

df %>% 
  filter(is.na(release_clause_eur)) %$%
  kable(head(.),"html") %>% 
  kable_styling(bootstrap_options = c("striped","hover", "condensed"), font_size = 8) %>% 
  scroll_box(width = "100%")

sofifa_id	player_url	short_name	long_name	player_positions	overall	potential	value_eur	wage_eur	age	dob	height_cm	weight_kg	club_team_id	club_name	league_name	league_level	club_position	club_jersey_number	club_loaned_from	club_contract_valid_until	nationality_id	nationality_name	nation_team_id	nation_position	nation_jersey_number	preferred_foot	weak_foot	skill_moves	international_reputation	work_rate	body_type	real_face	release_clause_eur	player_tags	player_traits	pace	shooting	passing	dribbling	defending	physic	attacking_crossing	attacking_finishing	attacking_heading_accuracy	attacking_short_passing	attacking_volleys	skill_dribbling	skill_curve	skill_fk_accuracy	skill_long_passing	skill_ball_control	movement_acceleration	movement_sprint_speed	movement_agility	movement_reactions	movement_balance	power_shot_power	power_jumping	power_stamina	power_strength	power_long_shots	mentality_aggression	mentality_interceptions	mentality_positioning	mentality_vision	mentality_penalties	mentality_composure	defending_marking_awareness	defending_standing_tackle	defending_sliding_tackle	goalkeeping_diving	goalkeeping_handling	goalkeeping_kicking	goalkeeping_positioning	goalkeeping_reflexes	goalkeeping_speed	ls	st	rs	lw	lf	cf	rf	rw	lam	cam	ram	lm	lcm	cm	rcm	rm	lwb	ldm	cdm	rdm	rwb	lb	lcb	cb	rcb	rb	gk	player_face_url	club_logo_url	club_flag_url	nation_logo_url	nation_flag_url
194765	https://sofifa.com/player/194765/antoine-griezmann/220002	A. Griezmann	Antoine Griezmann	ST, LW, RW	85	85	53000000	220000	30	1991-03-21	176	73	240	Atlético de Madrid	Spain Primera Division	1	LS	17	FC Barcelona	2022	18	France	1335	RW	7	Left	3	4	4	Medium/Medium	Unique	Yes	NA	#Acrobat	Finesse Shot, Flair, Long Shot Taker (AI), Playmaker (AI), Outside Foot Shot, Chip Shot (AI), Technical Dribbler (AI)	80	84	84	87	52	72	83	84	83	84	86	85	86	85	82	89	80	80	92	89	83	82	90	86	63	83	73	49	89	85	79	90	43	54	49	14	8	14	13	14	NA	84+1	84+1	84+1	85	86-1	86-1	86-1	85	86-1	86-1	86-1	85	82+3	82+3	82+3	85	71+3	69+3	69+3	69+3	71+3	69+3	63+3	63+3	63+3	69+3	21+3	https://cdn.sofifa.net/players/194/765/22_120.png	https://cdn.sofifa.net/teams/240/60.png	https://cdn.sofifa.net/flags/es.png	https://cdn.sofifa.net/teams/1335/60.png	https://cdn.sofifa.net/flags/fr.png
184087	https://sofifa.com/player/184087/toby-alderweireld/220002	T. Alderweireld	Toby Alderweireld	CB	83	83	0	0	32	1989-03-02	186	81	0			6		0		0	7	Belgium	1325	RCB	2	Right	3	2	3	Medium/Medium	Normal (185+)	Yes	NA		Long Passer (AI)	58	55	70	67	86	77	64	45	81	77	38	62	63	59	81	75	55	60	54	85	62	78	81	76	77	58	79	85	52	62	58	86	87	87	84	16	6	14	16	14	NA	65+3	65+3	65+3	63	65	65	65	63	66+3	66+3	66+3	66+3	73+3	73+3	73+3	66+3	77+3	81+2	81+2	81+2	77+3	78+3	82+1	82+1	82+1	78+3	21+3	https://cdn.sofifa.net/players/184/087/22_120.png			https://cdn.sofifa.net/teams/1325/60.png	https://cdn.sofifa.net/flags/be.png
201153	https://sofifa.com/player/201153/alvaro-borja-morata-martin/220002	Morata	Álvaro Borja Morata Martín	ST	83	83	37000000	84000	28	1992-10-23	190	84	45	Juventus	Italian Serie A	1	ST	9	Atlético de Madrid	2022	45	Spain	1362	SUB	7	Right	4	3	3	Medium/Medium	Unique	Yes	NA		Speed Dribbler (AI)	82	80	72	81	31	77	72	84	86	78	80	83	78	44	60	83	79	85	72	80	63	80	84	77	80	72	69	24	87	77	75	79	37	14	20	4	5	4	4	5	NA	83	83	83	81	82	82	82	81	80+3	80+3	80+3	79+3	71+3	71+3	71+3	79+3	58+3	55+3	55+3	55+3	58+3	54+3	50+3	50+3	50+3	54+3	13+3	https://cdn.sofifa.net/players/201/153/22_120.png	https://cdn.sofifa.net/teams/45/60.png	https://cdn.sofifa.net/flags/it.png	https://cdn.sofifa.net/teams/1362/60.png	https://cdn.sofifa.net/flags/es.png
235805	https://sofifa.com/player/235805/federico-chiesa/220002	F. Chiesa	Federico Chiesa	RW, LW, RM	83	91	80500000	74000	23	1997-10-25	175	70	45	Juventus	Italian Serie A	1	LM	22	Fiorentina	2022	27	Italy	1343	RW	14	Right	4	4	3	High/Medium	Normal (170-185)	Yes	NA	#Speedster, #Dribbler, #Acrobat	Long Shot Taker (AI), Speed Dribbler (AI)	91	81	74	85	48	73	73	79	50	78	80	89	78	52	72	82	91	91	87	83	81	86	53	85	71	84	69	30	81	75	62	78	65	44	44	6	7	8	9	7	NA	79+3	79+3	79+3	83	82	82	82	83	82+3	82+3	82+3	82+3	75+3	75+3	75+3	82+3	67+3	64+3	64+3	64+3	67+3	64+3	57+3	57+3	57+3	64+3	16+3	https://cdn.sofifa.net/players/235/805/22_120.png	https://cdn.sofifa.net/teams/45/60.png	https://cdn.sofifa.net/flags/it.png	https://cdn.sofifa.net/teams/1343/60.png	https://cdn.sofifa.net/flags/it.png
180206	https://sofifa.com/player/180206/miralem-pjanic/220002	M. Pjanić	Miralem Pjanić	CM	82	82	25000000	155000	31	1990-04-02	178	72	327	Beşiktaş JK	Turkish Süper Lig	1	RCM	15	FC Barcelona	2022	8	Bosnia and Herzegovina	NA		NA	Right	4	3	3	Medium/Medium	Normal (170-185)	Yes	NA	#FK Specialist	Finesse Shot, Playmaker (AI), Outside Foot Shot, Technical Dribbler (AI)	65	68	83	81	75	67	80	56	60	84	72	81	86	92	82	83	67	64	74	83	80	78	59	79	60	80	70	78	68	84	79	84	78	77	71	7	7	13	7	8	NA	70+3	70+3	70+3	76	75	75	75	76	78+3	78+3	78+3	77+3	81+1	81+1	81+1	77+3	78+3	79+3	79+3	79+3	78+3	76+3	73+3	73+3	73+3	76+3	16+3	https://cdn.sofifa.net/players/180/206/22_120.png	https://cdn.sofifa.net/teams/327/60.png	https://cdn.sofifa.net/flags/tr.png		https://cdn.sofifa.net/flags/ba.png
193105	https://sofifa.com/player/193105/alphonse-areola/220002	A. Areola	Alphonse Areola	GK	82	84	26000000	75000	28	1993-02-27	195	94	19	West Ham United	English Premier League	1	SUB	13	Paris Saint-Germain	2022	18	France	NA		NA	Right	3	1	2	Medium/Medium	Unique	Yes	NA		Comes For Crosses	NA	NA	NA	NA	NA	NA	20	19	14	48	16	15	16	16	37	22	56	54	58	78	58	57	72	38	80	14	26	23	17	51	25	64	13	18	12	85	79	76	80	85	55	34+2	34+2	34+2	33	34	34	34	33	36+2	36+2	36+2	35+2	36+2	36+2	36+2	35+2	32+2	34+2	34+2	34+2	32+2	31+2	31+2	31+2	31+2	31+2	81+2	https://cdn.sofifa.net/players/193/105/22_120.png	https://cdn.sofifa.net/teams/19/60.png	https://cdn.sofifa.net/flags/gb-eng.png		https://cdn.sofifa.net/flags/fr.png

The player with missing release_clause_eur are player whom didn’t have release clause so it equal to 0.

df$release_clause_eur[is.na(df$release_clause_eur)] <- 0

plot_missing(df, missing_only = TRUE, title = "Percentage of Missing Features")

Goalkeeping player

From the data we can see that there are 11.08% of player which missing basic features such as physic, defending, dribbling, passing, shooting, and pace. On the other hand, there are 88.92% of player who missing goalkeeping_speed. If we combine these two numbers we will got 100%. Therefore, There might be two group of player Goalkeeper and Non-Goalkeeper.

# Check non-goalkeeper group
df %>% 
  filter(is.na(goalkeeping_speed)) %$%
  unique(club_position)

##  [1] "RW"  "ST"  "LW"  "RCM" "CF"  "CDM" "LCB" "RDM" "RS"  "LCM" "SUB" "CAM"
## [13] "RCB" "LDM" "LB"  "RB"  "LM"  "RM"  "LS"  "CB"  "RES" ""    "RWB" "RF" 
## [25] "CM"  "LWB" "LAM" "LF"  "RAM"

As we expect these group of players are Non-Goalkeeper. we can fill goalkeeping_speed with 0 since it not related with their position.

df$goalkeeping_speed[is.na(df$goalkeeping_speed)] <- 0

# Check goalkeeper group
df %>% 
  filter(is.na(physic) &
         is.na(defending) &
         is.na(dribbling) &
         is.na(passing) &
         is.na(shooting)& 
         is.na(pace)) %$% 
  unique(club_position)

## [1] "GK"  "SUB" "RES" ""

Apparently there are not only GK in these group need keep investigating.

df %>% 
  filter(is.na(physic) &
         is.na(defending) &
         is.na(dribbling) &
         is.na(passing) &
         is.na(shooting)& 
         is.na(pace)) %>% 
  filter(club_position %in% c("SUB", "RES", "")) %$%
  unique(player_positions)

## [1] "GK"

After more investigation the player with club_position equal to SUB, RES, and "" in second group also a Goalkeeper. and we can fill these features with 0.

df$physic[is.na(df$physic)] <- 0
df$defending[is.na(df$defending)] <- 0
df$dribbling[is.na(df$dribbling)] <- 0
df$passing[is.na(df$passing)] <- 0
df$shooting[is.na(df$shooting)] <-0
df$pace[is.na(df$pace)] <- 0

plot_missing(df, missing_only = TRUE, title = "Percentage of Missing Features")

National Team player

The missing value of nation_jersey_number and nation_team_id is a majority of player who not selected in their national team. we will fill this with 0.

df$nation_jersey_number[is.na(df$nation_jersey_number)] <- 0
df$nation_team_id[is.na(df$nation_team_id)] <- 0

plot_intro(df, title = "Data Structure of database")

With this we deal with all missing value and reach tidydata.

# save df into .Rdata format
save(df, file = "../output/fifa_22_tidydata_raw.Rdata")
list.files(path = "../output")

## [1] "FIFA_22_Analysis_files"         "FIFA_22_Analysis.html"         
## [3] "fifa_22_tidydata_cleaned.Rdata" "fifa_22_tidydata_raw.Rdata"

Data Manipulation

Unnecessary Features

df %<>% select(-player_url, -long_name, -real_face, -ls, -st, -rs, -lw, -lf, -cf, -rf, -rw, -lam, -cam, -ram, -lm, -lcm, -cm, -rcm, -rm, -lwb, -ldm, -cdm, -rdm, -rwb, -lb, -lcb, -cb, -rcb, -rb, -gk)
kable(t(head(df)), "html") %>% kable_styling(bootstrap_options = c("striped","hover", "condensed"), font_size = 8) %>% scroll_box(width = "100%")

	1	2	3	4	5	6
sofifa_id	158023	188545	20801	190871	192985	200389
short_name	L. Messi	R. Lewandowski	Cristiano Ronaldo	Neymar Jr	K. De Bruyne	J. Oblak
player_positions	RW, ST, CF	ST	ST, LW	LW, CAM	CM, CAM	GK
overall	93	92	91	91	91	91
potential	93	92	91	91	91	93
value_eur	78000000	119500000	45000000	129000000	125500000	112000000
wage_eur	320000	270000	270000	270000	350000	130000
age	34	32	36	29	30	28
dob	1987-06-24	1988-08-21	1985-02-05	1992-02-05	1991-06-28	1993-01-07
height_cm	170	185	187	175	181	188
weight_kg	72	81	83	68	70	87
club_team_id	73	21	11	73	10	240
club_name	Paris Saint-Germain	FC Bayern München	Manchester United	Paris Saint-Germain	Manchester City	Atlético de Madrid
league_name	French Ligue 1	German 1. Bundesliga	English Premier League	French Ligue 1	English Premier League	Spain Primera Division
league_level	1	1	1	1	1	1
club_position	RW	ST	ST	LW	RCM	GK
club_jersey_number	30	9	7	10	17	13
club_loaned_from
club_joined	2021-08-10	2014-07-01	2021-08-27	2017-08-03	2015-08-30	2014-07-16
club_contract_valid_until	2023	2023	2023	2025	2025	2023
nationality_id	52	37	38	54	7	44
nationality_name	Argentina	Poland	Portugal	Brazil	Belgium	Slovenia
nation_team_id	1369	1353	1354	0	1325	0
nation_position	RW	RS	ST		RCM
nation_jersey_number	10	9	7	0	7	0
preferred_foot	Left	Right	Right	Right	Right	Right
weak_foot	4	4	4	5	5	3
skill_moves	4	4	5	5	4	1
international_reputation	5	5	5	5	4	5
work_rate	Medium/Low	High/Medium	High/Low	High/Medium	High/High	Medium/Medium
body_type	Unique	Unique	Unique	Unique	Unique	Unique
release_clause_eur	144300000	197200000	83300000	238700000	232200000	238000000
player_tags	#Dribbler, #Distance Shooter, #FK Specialist, #Acrobat, #Clinical Finisher, #Complete Forward	#Aerial Threat, #Distance Shooter, #Clinical Finisher, #Complete Forward	#Aerial Threat, #Dribbler, #Distance Shooter, #Crosser, #Acrobat, #Clinical Finisher, #Complete Forward	#Speedster, #Dribbler, #Playmaker, #FK Specialist, #Acrobat, #Complete Midfielder	#Dribbler, #Playmaker, #Engine, #Distance Shooter, #Crosser, #Complete Midfielder
player_traits	Finesse Shot, Long Shot Taker (AI), Playmaker (AI), Outside Foot Shot, One Club Player, Chip Shot (AI), Technical Dribbler (AI)	Solid Player, Finesse Shot, Outside Foot Shot, Chip Shot (AI)	Power Free-Kick, Flair, Long Shot Taker (AI), Speed Dribbler (AI), Outside Foot Shot	Injury Prone, Flair, Speed Dribbler (AI), Playmaker (AI), Outside Foot Shot, Technical Dribbler (AI)	Injury Prone, Leadership, Early Crosser, Long Passer (AI), Long Shot Taker (AI), Playmaker (AI), Outside Foot Shot	GK Long Throw, Comes For Crosses
pace	85	78	87	91	76	0
shooting	92	92	94	83	86	0
passing	91	79	80	86	93	0
dribbling	95	86	88	94	88	0
defending	34	44	34	37	64	0
physic	65	82	75	63	78	0
attacking_crossing	85	71	87	85	94	13
attacking_finishing	95	95	95	83	82	11
attacking_heading_accuracy	70	90	90	63	55	15
attacking_short_passing	91	85	80	86	94	43
attacking_volleys	88	89	86	86	82	13
skill_dribbling	96	85	88	95	88	12
skill_curve	93	79	81	88	85	13
skill_fk_accuracy	94	85	84	87	83	14
skill_long_passing	91	70	77	81	93	40
skill_ball_control	96	88	88	95	91	30
movement_acceleration	91	77	85	93	76	43
movement_sprint_speed	80	79	88	89	76	60
movement_agility	91	77	86	96	79	67
movement_reactions	94	93	94	89	91	88
movement_balance	95	82	74	84	78	49
power_shot_power	86	90	94	80	91	59
power_jumping	68	85	95	64	63	78
power_stamina	72	76	77	81	89	41
power_strength	69	86	77	53	74	78
power_long_shots	94	87	93	81	91	12
mentality_aggression	44	81	63	63	76	34
mentality_interceptions	40	49	29	37	66	19
mentality_positioning	93	95	95	86	88	11
mentality_vision	95	81	76	90	94	65
mentality_penalties	75	90	88	93	83	11
mentality_composure	96	88	95	93	89	68
defending_marking_awareness	20	35	24	35	68	27
defending_standing_tackle	35	42	32	32	65	12
defending_sliding_tackle	24	19	24	29	53	18
goalkeeping_diving	6	15	7	9	15	87
goalkeeping_handling	11	6	11	9	13	92
goalkeeping_kicking	15	12	15	15	5	78
goalkeeping_positioning	14	8	14	15	10	90
goalkeeping_reflexes	8	10	11	11	13	90
goalkeeping_speed	0	0	0	0	0	50
player_face_url	https://cdn.sofifa.net/players/158/023/22_120.png	https://cdn.sofifa.net/players/188/545/22_120.png	https://cdn.sofifa.net/players/020/801/22_120.png	https://cdn.sofifa.net/players/190/871/22_120.png	https://cdn.sofifa.net/players/192/985/22_120.png	https://cdn.sofifa.net/players/200/389/22_120.png
club_logo_url	https://cdn.sofifa.net/teams/73/60.png	https://cdn.sofifa.net/teams/21/60.png	https://cdn.sofifa.net/teams/11/60.png	https://cdn.sofifa.net/teams/73/60.png	https://cdn.sofifa.net/teams/10/60.png	https://cdn.sofifa.net/teams/240/60.png
club_flag_url	https://cdn.sofifa.net/flags/fr.png	https://cdn.sofifa.net/flags/de.png	https://cdn.sofifa.net/flags/gb-eng.png	https://cdn.sofifa.net/flags/fr.png	https://cdn.sofifa.net/flags/gb-eng.png	https://cdn.sofifa.net/flags/es.png
nation_logo_url	https://cdn.sofifa.net/teams/1369/60.png	https://cdn.sofifa.net/teams/1353/60.png	https://cdn.sofifa.net/teams/1354/60.png		https://cdn.sofifa.net/teams/1325/60.png
nation_flag_url	https://cdn.sofifa.net/flags/ar.png	https://cdn.sofifa.net/flags/pl.png	https://cdn.sofifa.net/flags/pt.png	https://cdn.sofifa.net/flags/br.png	https://cdn.sofifa.net/flags/be.png	https://cdn.sofifa.net/flags/si.png

Stacked columns

work_rate, player_tags, and player_traits are stacked columns. This mean in each columns there might be more than one values in it.

# Unstacked work_rate

df <- df %$%
  cSplit(., 'work_rate', sep= "/", type.convert = FALSE) %>% 
  rename(
    work_rate_att = work_rate_1,
    work_rate_dff = work_rate_2 )

kable(t(head(df,3)), "html") %>% kable_styling(bootstrap_options = c("striped","hover", "condensed"), font_size = 8) %>% scroll_box(width = "100%")

sofifa_id	158023	188545	20801
short_name	L. Messi	R. Lewandowski	Cristiano Ronaldo
player_positions	RW, ST, CF	ST	ST, LW
overall	93	92	91
potential	93	92	91
value_eur	78000000	119500000	45000000
wage_eur	320000	270000	270000
age	34	32	36
dob	1987-06-24	1988-08-21	1985-02-05
height_cm	170	185	187
weight_kg	72	81	83
club_team_id	73	21	11
club_name	Paris Saint-Germain	FC Bayern München	Manchester United
league_name	French Ligue 1	German 1. Bundesliga	English Premier League
league_level	1	1	1
club_position	RW	ST	ST
club_jersey_number	30	9	7
club_loaned_from
club_joined	2021-08-10	2014-07-01	2021-08-27
club_contract_valid_until	2023	2023	2023
nationality_id	52	37	38
nationality_name	Argentina	Poland	Portugal
nation_team_id	1369	1353	1354
nation_position	RW	RS	ST
nation_jersey_number	10	9	7
preferred_foot	Left	Right	Right
weak_foot	4	4	4
skill_moves	4	4	5
international_reputation	5	5	5
body_type	Unique	Unique	Unique
release_clause_eur	144300000	197200000	83300000
player_tags	#Dribbler, #Distance Shooter, #FK Specialist, #Acrobat, #Clinical Finisher, #Complete Forward	#Aerial Threat, #Distance Shooter, #Clinical Finisher, #Complete Forward	#Aerial Threat, #Dribbler, #Distance Shooter, #Crosser, #Acrobat, #Clinical Finisher, #Complete Forward
player_traits	Finesse Shot, Long Shot Taker (AI), Playmaker (AI), Outside Foot Shot, One Club Player, Chip Shot (AI), Technical Dribbler (AI)	Solid Player, Finesse Shot, Outside Foot Shot, Chip Shot (AI)	Power Free-Kick, Flair, Long Shot Taker (AI), Speed Dribbler (AI), Outside Foot Shot
pace	85	78	87
shooting	92	92	94
passing	91	79	80
dribbling	95	86	88
defending	34	44	34
physic	65	82	75
attacking_crossing	85	71	87
attacking_finishing	95	95	95
attacking_heading_accuracy	70	90	90
attacking_short_passing	91	85	80
attacking_volleys	88	89	86
skill_dribbling	96	85	88
skill_curve	93	79	81
skill_fk_accuracy	94	85	84
skill_long_passing	91	70	77
skill_ball_control	96	88	88
movement_acceleration	91	77	85
movement_sprint_speed	80	79	88
movement_agility	91	77	86
movement_reactions	94	93	94
movement_balance	95	82	74
power_shot_power	86	90	94
power_jumping	68	85	95
power_stamina	72	76	77
power_strength	69	86	77
power_long_shots	94	87	93
mentality_aggression	44	81	63
mentality_interceptions	40	49	29
mentality_positioning	93	95	95
mentality_vision	95	81	76
mentality_penalties	75	90	88
mentality_composure	96	88	95
defending_marking_awareness	20	35	24
defending_standing_tackle	35	42	32
defending_sliding_tackle	24	19	24
goalkeeping_diving	6	15	7
goalkeeping_handling	11	6	11
goalkeeping_kicking	15	12	15
goalkeeping_positioning	14	8	14
goalkeeping_reflexes	8	10	11
goalkeeping_speed	0	0	0
player_face_url	https://cdn.sofifa.net/players/158/023/22_120.png	https://cdn.sofifa.net/players/188/545/22_120.png	https://cdn.sofifa.net/players/020/801/22_120.png
club_logo_url	https://cdn.sofifa.net/teams/73/60.png	https://cdn.sofifa.net/teams/21/60.png	https://cdn.sofifa.net/teams/11/60.png
club_flag_url	https://cdn.sofifa.net/flags/fr.png	https://cdn.sofifa.net/flags/de.png	https://cdn.sofifa.net/flags/gb-eng.png
nation_logo_url	https://cdn.sofifa.net/teams/1369/60.png	https://cdn.sofifa.net/teams/1353/60.png	https://cdn.sofifa.net/teams/1354/60.png
nation_flag_url	https://cdn.sofifa.net/flags/ar.png	https://cdn.sofifa.net/flags/pl.png	https://cdn.sofifa.net/flags/pt.png
work_rate_att	Medium	High	High
work_rate_dff	Low	Medium	Low

for feature like player_tags, and player_traits we need to one-hot encode them.

one_hot <- function(x) {
  map_df(x, table) %>% 
    mutate_all(as.integer) %>% 
    mutate_all(replace_na, 0L)
}
df$player_traits <- str_remove_all(df$player_traits, "\\(AI\\)")

df$player_tags[df$player_tags == ""] <- "#No_Tags"
df$player_traits[df$player_traits == ""] <- "No Traits"


df %<>% 
  mutate(one_hot(strsplit(player_tags, ", "))) %>% 
  mutate(one_hot(strsplit(player_traits, ", "))) %>% 
  select(-player_tags, -player_traits) 

df %$% 
  kable(t(head(.,3)), "html") %>% 
  kable_styling(bootstrap_options = c("striped","hover", "condensed"), font_size = 8) %>% 
  scroll_box(width = "100%")

sofifa_id	158023	188545	20801
short_name	L. Messi	R. Lewandowski	Cristiano Ronaldo
player_positions	RW, ST, CF	ST	ST, LW
overall	93	92	91
potential	93	92	91
value_eur	78000000	119500000	45000000
wage_eur	320000	270000	270000
age	34	32	36
dob	1987-06-24	1988-08-21	1985-02-05
height_cm	170	185	187
weight_kg	72	81	83
club_team_id	73	21	11
club_name	Paris Saint-Germain	FC Bayern München	Manchester United
league_name	French Ligue 1	German 1. Bundesliga	English Premier League
league_level	1	1	1
club_position	RW	ST	ST
club_jersey_number	30	9	7
club_loaned_from
club_joined	2021-08-10	2014-07-01	2021-08-27
club_contract_valid_until	2023	2023	2023
nationality_id	52	37	38
nationality_name	Argentina	Poland	Portugal
nation_team_id	1369	1353	1354
nation_position	RW	RS	ST
nation_jersey_number	10	9	7
preferred_foot	Left	Right	Right
weak_foot	4	4	4
skill_moves	4	4	5
international_reputation	5	5	5
body_type	Unique	Unique	Unique
release_clause_eur	144300000	197200000	83300000
pace	85	78	87
shooting	92	92	94
passing	91	79	80
dribbling	95	86	88
defending	34	44	34
physic	65	82	75
attacking_crossing	85	71	87
attacking_finishing	95	95	95
attacking_heading_accuracy	70	90	90
attacking_short_passing	91	85	80
attacking_volleys	88	89	86
skill_dribbling	96	85	88
skill_curve	93	79	81
skill_fk_accuracy	94	85	84
skill_long_passing	91	70	77
skill_ball_control	96	88	88
movement_acceleration	91	77	85
movement_sprint_speed	80	79	88
movement_agility	91	77	86
movement_reactions	94	93	94
movement_balance	95	82	74
power_shot_power	86	90	94
power_jumping	68	85	95
power_stamina	72	76	77
power_strength	69	86	77
power_long_shots	94	87	93
mentality_aggression	44	81	63
mentality_interceptions	40	49	29
mentality_positioning	93	95	95
mentality_vision	95	81	76
mentality_penalties	75	90	88
mentality_composure	96	88	95
defending_marking_awareness	20	35	24
defending_standing_tackle	35	42	32
defending_sliding_tackle	24	19	24
goalkeeping_diving	6	15	7
goalkeeping_handling	11	6	11
goalkeeping_kicking	15	12	15
goalkeeping_positioning	14	8	14
goalkeeping_reflexes	8	10	11
goalkeeping_speed	0	0	0
player_face_url	https://cdn.sofifa.net/players/158/023/22_120.png	https://cdn.sofifa.net/players/188/545/22_120.png	https://cdn.sofifa.net/players/020/801/22_120.png
club_logo_url	https://cdn.sofifa.net/teams/73/60.png	https://cdn.sofifa.net/teams/21/60.png	https://cdn.sofifa.net/teams/11/60.png
club_flag_url	https://cdn.sofifa.net/flags/fr.png	https://cdn.sofifa.net/flags/de.png	https://cdn.sofifa.net/flags/gb-eng.png
nation_logo_url	https://cdn.sofifa.net/teams/1369/60.png	https://cdn.sofifa.net/teams/1353/60.png	https://cdn.sofifa.net/teams/1354/60.png
nation_flag_url	https://cdn.sofifa.net/flags/ar.png	https://cdn.sofifa.net/flags/pl.png	https://cdn.sofifa.net/flags/pt.png
work_rate_att	Medium	High	High
work_rate_dff	Low	Medium	Low
#Acrobat	1	0	1
#Clinical Finisher	1	1	1
#Complete Forward	1	1	1
#Distance Shooter	1	1	1
#Dribbler	1	0	1
#FK Specialist	1	0	0
#Aerial Threat	0	1	1
#Crosser	0	0	1
#Complete Midfielder	0	0	0
#Playmaker	0	0	0
#Speedster	0	0	0
#Engine	0	0	0
#No_Tags	0	0	0
#Tackling	0	0	0
#Tactician	0	0	0
#Poacher	0	0	0
#Complete Defender	0	0	0
#Strength	0	0	0
#Tactician	0	0	0
#Tackling	0	0	0
#Playmaker	0	0	0
Chip Shot	1	1	0
Finesse Shot	1	1	0
Long Shot Taker	1	0	1
One Club Player	1	0	0
Outside Foot Shot	1	1	1
Playmaker	1	0	0
Technical Dribbler	1	0	0
Solid Player	0	1	0
Flair	0	0	1
Power Free-Kick	0	0	1
Speed Dribbler	0	0	1
Injury Prone	0	0	0
Early Crosser	0	0	0
Leadership	0	0	0
Long Passer	0	0	0
Comes For Crosses	0	0	0
GK Long Throw	0	0	0
Rushes Out Of Goal	0	0	0
Saves with Feet	0	0	0
Team Player	0	0	0
Dives Into Tackles	0	0	0
Power Header	0	0	0
Cautious With Crosses	0	0	0
Long Throw-in	0	0	0
No Traits	0	0	0
Giant Throw-in	0	0	0

one player can plays multiple position. However, as we can see the first postion in player_postion is there main position. Therefore, we are going to make it their only position.

df$player_positions <- sapply(strsplit(df$player_positions, ", "), `[`, 1)

defence <- c("CB", "RB", "LB", "LWB", "RWB", "LCB", "RCB")
midfielder <- c("CM", "CDM","CAM","LM","RM", "LAM", "RAM", "LCM", "RCM", "LDM", "RDM")

df %<>% mutate(Class = if_else(player_positions %in% "GK", "GK",
                                 if_else(player_positions %in% defence, "DEF",
                                         if_else(player_positions %in% midfielder, "MID", "FWD"))))

rm(defence, midfielder)

Now we finished cleaning our data and can begins ours analysis.

# save df into .Rdata format
save(df, file = "../output/fifa_22_tidydata_cleaned.Rdata")
list.files(path = "../output")

## [1] "FIFA_22_Analysis_files"         "FIFA_22_Analysis.html"         
## [3] "fifa_22_tidydata_cleaned.Rdata" "fifa_22_tidydata_raw.Rdata"

EDA

load("../output/fifa_22_tidydata_cleaned.Rdata")

df %>% 
  group_by(nationality_name) %>% 
  summarise(n_player = n()) %>% 
  arrange(desc(n_player)) %$%
  kable((head(.,10)), "html") %>%
  kable_styling(bootstrap_options = c("striped","hover", "condensed"), font_size = 8) %>%
  scroll_box(width = "100%")

nationality_name	n_player
England	1719
Germany	1214
Spain	1086
France	980
Argentina	960
Brazil	897
Japan	546
Netherlands	439
United States	413
Poland	403

options(repr.plot.width = 12, repr.plot.heigh = 8)

world_map <- map_data("world")

numofplayers <- world_map %>% 
  mutate(region = as.character(region)) %>% 
  left_join((df %>% mutate(Nationality = as.character(nationality_name),
                           Nationality = if_else(nationality_name %in% "England", 
                                                 "UK", nationality_name)) %>%
               count(Nationality, name = "Number of Player") %>%
               rename(region = Nationality) %>%
               mutate(region = as.character(region))), by = "region")

ggplotly(ggplot(numofplayers, aes(long, lat, group = group))+
    geom_polygon(aes(fill = `Number of Player` ), color = "white", show.legend = FALSE)+
    scale_fill_viridis_c(option = "C")+
    theme_fivethirtyeight()+
    labs(fill = "Number of Player",
         title = "Number of Player From Around the World"))

There are 163 countries in the database and European Countries have most player. The England has the most player in the game with 1719 players, following with Germany and Spain with 1214 and 1086 players.

n_clubs <- length(unique(df$club_team_id))
paste0("Total number of clubs :", n_clubs)

## [1] "Total number of clubs :702"

max_pa <- df %>% 
  arrange(desc(potential)) %>% 
  slice(1:1) 

max_ca <- df %>% 
  arrange(desc(overall)) %>% 
  slice(1:1) 

cat("Best Player\n--------------------------------------------\n",paste0("Maximum Potentia :", max_pa$short_name),"\n",paste0("Maximum Overall Perforamnce :", max_ca$short_name))

## Best Player
## --------------------------------------------
##  Maximum Potentia :K. Mbappé 
##  Maximum Overall Perforamnce :L. Messi

atts_list <- c("pace", "shooting", "passing", "dribbling", "defending", "physic", "attacking_crossing", "attacking_finishing", "attacking_heading_accuracy", "attacking_short_passing", "attacking_volleys", "skill_dribbling", "skill_curve", "skill_fk_accuracy",  "skill_long_passing", "skill_ball_control", "movement_acceleration", "movement_sprint_speed", "movement_agility", "movement_reactions", "movement_balance", "power_shot_power", "power_jumping", "power_stamina", "power_strength", "power_long_shots", "mentality_aggression", "mentality_interceptions", "mentality_positioning", "mentality_vision", "mentality_penalties","mentality_composure", "defending_marking_awareness", "defending_standing_tackle", "defending_sliding_tackle", "goalkeeping_diving", "goalkeeping_handling", "goalkeeping_kicking", "goalkeeping_positioning", "goalkeeping_reflexes", "goalkeeping_speed")

atts <- df %>% 
  select(short_name, all_of(atts_list)) 
  
bestof = melt(atts, id.vars = "short_name", variable.name = "Attribute")
bestof = bestof[, .SD[which.max(value)], by= Attribute][, c("Attribute", "short_name")]

kable(bestof, "html") %>%
  kable_styling(bootstrap_options = c("striped","hover", "condensed"), font_size = 8) %>% 
  scroll_box(width = "100%")

Attribute	short_name
pace	K. Mbappé
shooting	Cristiano Ronaldo
passing	K. De Bruyne
dribbling	L. Messi
defending	V. van Dijk
physic	Casemiro
attacking_crossing	K. De Bruyne
attacking_finishing	L. Messi
attacking_heading_accuracy	L. de Jong
attacking_short_passing	K. De Bruyne
attacking_volleys	L. Suárez
skill_dribbling	L. Messi
skill_curve	Quaresma
skill_fk_accuracy	L. Messi
skill_long_passing	K. De Bruyne
skill_ball_control	L. Messi
movement_acceleration	K. Mbappé
movement_sprint_speed	K. Mbappé
movement_agility	Neymar Jr
movement_reactions	L. Messi
movement_balance	R. Fraser
power_shot_power	A. Kolarov
power_jumping	Cristiano Ronaldo
power_stamina	N. Kanté
power_strength	A. Akinfenwa
power_long_shots	L. Messi
mentality_aggression	B. Pearson
mentality_interceptions	N. Kanté
mentality_positioning	T. Müller
mentality_vision	L. Messi
mentality_penalties	Neymar Jr
mentality_composure	L. Messi
defending_marking_awareness	G. Chiellini
defending_standing_tackle	N. Kanté
defending_sliding_tackle	A. Wan-Bissaka
goalkeeping_diving	G. Donnarumma
goalkeeping_handling	J. Oblak
goalkeeping_kicking	Ederson
goalkeeping_positioning	S. Handanovič
goalkeeping_reflexes	J. Oblak
goalkeeping_speed	Jordi Masip

max_value <- df %>%
  arrange(desc(value_eur)) %>% 
  slice(1:1)

max_wage <- df %>% 
  arrange(desc(wage_eur)) %>% 
  slice(1:1)
cat("Top Earners\n----------------------------------\n", paste0("Maximum Value :", max_value$short_name), "\n", paste0("Maximum Wages :", max_wage$short_name))

## Top Earners
## ----------------------------------
##  Maximum Value :K. Mbappé 
##  Maximum Wages :K. De Bruyne

foots <- df %>% 
  group_by(preferred_foot) %>% 
  summarise(n_player = n()) 

foots$p_player <- round(100 * (foots$n_player / sum(foots$n_player)))


ggplot(foots, aes(fill=preferred_foot, values=p_player))+
  geom_waffle(n_rows= 10, colour= "White") +
  scale_fill_manual(name= NULL,
                    values = c("grey50", "grey"),
                    labels= c("Left", "Right")) +
  ggtitle("Left vs Right foot", subtitle = "There are 24 lefty out of every 100 players") +
  coord_equal() +
  theme_fivethirtyeight() +
  theme(panel.grid.major = element_blank(),
        axis.text.y = element_blank(), 
        axis.text.x = element_blank())

Player Overall

Player rating are normally distributed in FIFA 22, with a mean of 65.77 and standard deviation of 6.88.

df %>% 
  ggplot(aes(x = overall)) +
  geom_histogram(color= "white", fill="darkgrey") +
  ggtitle("Player Ratings Are Normally Distributed", subtitle = "The mean can be used as a measure of central tendancy") +
  theme_fivethirtyeight() +
  theme(axis.text.y = element_blank())

Now we can check the distibution of each leage.

top_league <- df %>% 
  filter(league_level == 1) %>% 
  group_by(league_name) %>% 
  summarise(avg_rating = mean(overall)) %>% 
  filter(avg_rating >= 70) %>% 
  select(league_name)

summ <- df %>% 
  filter(league_name %in% top_league$league_name) %>% 
  group_by(league_name) %>% 
  summarise(avg_rating = mean(overall),
            n_player = n(),
            avg_age = mean(age))

top_league_player <- df %>% 
  filter(league_name %in% top_league$league_name)

options(repr.plot.width = 12, repr.plot.heigh = 8)

ggplotly(ggplot() + 
  geom_histogram(top_league_player, mapping = aes(overall, fill= league_name), color= "white", fill="darkgrey") +
  geom_vline(summ, mapping = aes(xintercept = avg_rating), size = 0.5, color="red") +
  geom_text(summ, mapping = aes(x = avg_rating-5, y= 100, label= round(avg_rating, digits = 2))) +
  facet_wrap(league_name~.) +
  theme_fivethirtyeight() +
  theme(legend.position = "bottom", axis.text.y = element_blank()) +
  labs(y = element_blank(), x= "Overall Rating" , title = "Rating Distribution of Player in Top Leagues", subtitle = "Overall rating distribution and average overall rating of all top leagues"))

The average overall rating of each league is closely related with how skilled player in the league, the higher the average overall rating - the higher the skill. Apparently, player in Campeonato Brasileiro Série A, Czech Republic Gambrinus Liga, and Ukrainian Premier League from Brazil, Czech, and Ukrain are as skilled as player in German 1. Budesliga or French Ligue 1 from German and France and player in Spain Primera Division from Spain is the most skilled one.

Age vs Overall Rating

df %>% 
  filter(!player_positions == "GK") %>% 
  group_by(age) %>% 
  summarise(Potential = mean(potential),
            Overall = mean(overall)) %>% 
  ggplot(aes(x = age)) +
  geom_line(aes(y=Overall), color= "grey50", size=1)+
  ggtitle("Average rating is flaten over the years", subtitle = "Player ratings tend not to get better after the age of 30")+
  theme_fivethirtyeight()

As we can see the average overall rating is flaten over the ages and drop sharply after 40.

df %>% 
  filter(!Class == "GK") %>% 
  group_by(Class, age) %>% 
  summarise(Rating = mean(overall)) %>% 
  ggplot(aes(x= age, y= Rating, group= Class)) +
  geom_line(size = 1, color= "grey50") +
  xlim(15, 45) +
  ggtitle("Rating over the ages by position class") +
  theme_fivethirtyeight() +
  facet_wrap(~ Class, ncol=1) +
  theme(strip.background = element_rect(fill= "darkgrey"), strip.text = element_text(color= "white", face= "bold"))

With this relationship is explored by the major position groups, we can see that defender ratings tend to beging their decline earliest at around 33 years of age, while the decline starts somewhere closer to 35 for both attackers and midfielders.

When player are reaching their potential

df %>% 
  group_by(age) %>% 
  summarise(Potential = mean(potential),
            Overall = mean(overall)) %>% 
  ggplot(aes(x = age)) +
  geom_line(aes(y=Potential), color= "purple", size=1, linetype= "dashed")+
  geom_line(aes(y=Overall), color= "grey50", size=1)+
  annotate("text", x= 30, y=73, label="Potential meets overall\ntalent at 29 years old", color= "grey50") +
  ggtitle("Potential And Overall Talent Converges", subtitle = "The average ratings were taken for each age")+
  theme_fivethirtyeight()

It appear that the player potential and player over all converge around when they are 29.

gk_vars <- df %>% select(contains("goalkeeping")) %>% names()

spearman_cor_overall <- df %>% 
  filter(player_positions != "GK") %>% 
  select_if(is.numeric) %>% 
  select(-club_team_id, -league_level, -club_jersey_number, -club_contract_valid_until, -nationality_id, -nation_team_id, -nation_jersey_number, -all_of(gk_vars)) %>% 
  as.matrix() %>% 
  na.omit() %>% 
  cor(method = "spearman")

pearson_cor_overall <- df %>% 
  filter(player_positions != "GK") %>% 
  select_if(is.numeric) %>% 
  select(-club_team_id, -league_level, -club_jersey_number, -club_contract_valid_until, -nationality_id, -nation_team_id, -nation_jersey_number, -all_of(gk_vars)) %>% 
  as.matrix() %>% 
  na.omit() %>% 
  cor()

cor_colnames <- colnames(spearman_cor_overall)

spearman_cor_overall <- spearman_cor_overall[,2] %>% data.frame()

spearman_cor_overall <- cbind(cor_colnames, spearman_cor_overall) %>% arrange(desc(`.`))

pearson_cor_overall <- pearson_cor_overall[,2] %>% data.frame()

pearson_cor_overall <- cbind(cor_colnames, pearson_cor_overall) %>% arrange(desc(`.`))

spearman_cor_overall %>% left_join(pearson_cor_overall, by="cor_colnames") %>% rename(Feature = cor_colnames, Spearman = `..x`, Pearson = `..y`) %>% filter(Feature != "overall") %>% head(10) %$%
  kable((head(.,10)), "html") %>%
  kable_styling(bootstrap_options = c("striped","hover", "condensed"), font_size = 8) %>%
  scroll_box(width = "100%")

Feature	Spearman	Pearson
value_eur	0.8815896	0.5627151
movement_reactions	0.8721298	0.8755071
mentality_composure	0.7967710	0.8103614
attacking_short_passing	0.7694905	0.7799224
skill_ball_control	0.7675370	0.7636762
wage_eur	0.7624437	0.6056264
release_clause_eur	0.7228825	0.5366370
passing	0.6959918	0.7150010
dribbling	0.6790794	0.6664023
skill_dribbling	0.6149346	0.5723942

These are 10 highly correalated attrs with overall.

Which postion are skilled in what?

tile_data <- df %>% 
  select_if(is.numeric) %>% 
  select(-all_of(gk_vars)) %>% 
  left_join(df %>% select(sofifa_id, player_positions, Class), by= "sofifa_id") %>% 
  select(starts_with("attacking_"), starts_with("skill_"), starts_with("movement_"), starts_with("power_"), starts_with("mentality_"), starts_with("defending_"), player_positions, Class) %>% select(-skill_moves) %>% 
  rename_all(~stringr::str_replace_all(.,"^attacking_","")) %>%
  rename_all(~stringr::str_replace_all(.,"^skill_","")) %>%
  rename_all(~stringr::str_replace_all(.,"^movement_","")) %>%
  rename_all(~stringr::str_replace_all(.,"^power_","")) %>% 
  rename_all(~stringr::str_replace_all(.,"^mentality_",""))  %>% 
  rename_all(~stringr::str_replace_all(.,"^defending_",""))

tile_data %<>% filter(player_positions != "GK") %>% 
  gather(key= Attribute, value= Value, -player_positions, -Class) %>% 
  group_by(Class, player_positions, Attribute) %>% 
  summarise(MedianValue = median(Value, na.rm = T)) %>% 
  ggplot(aes(x= Attribute, y= player_positions)) +
  geom_tile(aes(fill = MedianValue), colour= "black") +
  geom_text(aes(label = MedianValue)) + 
  scale_fill_gradient(low= "purple", high= "green") +
  ggtitle("Defender are strong, Foward are agile", subtitle = "Analysing the median ratings for each of the attributes for each postion for player with and overall rating over 75") +
  theme_fivethirtyeight() +
  theme(axis.text.x = element_text(angle = 45, hjust= 1), strip.text = element_text(face= "bold", size=12), legend.position = "none") +
  facet_wrap(~ Class, scales = "free", ncol= 1)

Team Overall Talent

Top_20_clubs <- df %>% 
  group_by(club_name) %>% 
  summarise(AverageRating = mean(overall, na.rm= T)) %>% 
  arrange(desc(AverageRating)) %>% 
  head(n=20) %>% 
  pull(club_name)

df %>% 
  filter(club_name %in% Top_20_clubs) %>% 
  mutate(Top3 = ifelse(club_name %in% c("Juventus", "Paris Saint-Germain", "Inter"), "Yes", "No")) %>% 
  ggplot(aes(x= reorder(club_name, overall), y=overall, fill= Top3)) +
  geom_boxplot(color= "black") +
  scale_fill_manual(values = c("lightgrey", "purple")) +
  ggtitle("Juventus has the Hightest Overall", subtitle = "The average overall rating of the 20 highest rated teams in the game, sorted in decending order") +
  coord_flip() +
  theme_fivethirtyeight() +
  theme(legend.position = "none")

Fifa 22 Analysis

Jitrayu Punrattanapongs

1/20/2022